Following Cochrane review protocols to completion 10 years later: a retrospective cohort study and author survey

We analyzed patterns of publication of Cochrane review protocols (CRPs). We analyzed CRPs published in 2010, extracted their characteristics, and analyzed whether they were published by February 2018. We surveyed corresponding authors and Cochrane review groups to analyze reasons for nonpublication of protocols and analyzed factors predicting the time to publication. Of 576 CRPs, 446 (77.4%) were published as a full review and 130 (22.6%) were still unpublished in February 2018; among unpublished, 37 (28.5%) were withdrawn and 93 (71.5%) were still active. The most common authors' reason for abandoning a protocol was a lack of time to work on the review. The median time to publication was 2.78 years (range 0.96 to 8.05). Multivariate analysis showed that factors with the strongest association with shorter time to publication were review being an update and new authors added. Analysis only on methodological variables indicated that the strongest association for a shorter time until publication was found for including only published data. Almost a quarter of CRPs remains unpublished after 8 years. This figure is slightly higher than in a previous analysis 10 years ago. Strategies for enhancing completion of Cochrane reviews should be considered.